Cattle monitoring is an important demand in precision livestock and crossbreed quality control. Previous studies and products have been proposed to approach this problem, although several factors pose challenges for real-time data acquisition and analysis. In this work, we present a proof of concept prototype for a cattle crossbreed monitoring system based on wireless sensor networks. The hardware implementation, the sensor data acquisition system and the field tests are described in detail. Supervised machine learning algorithms are applied for copulation detection and the classification metrics show that some of the proposed models have good sensitivity, suggesting promising directions for future steps and optimization.